Input and Target

When you create a model, PerceptiLabs automatically creates and adds both an Input (1) and a Target (2) Component to your model:

The Input Component (1) represents the input data for your model and points to the .csv file you specified when you created the model. It provides a visualization using the first sample from your input data.

The Target Component (2) provides a visualization of the labels / the result we want the model to achieve (e.g., classification).

The screenshot above shows a basic image classification model, where the Input's visualization displays the first image of the input data while the Target's visualization shows the (normalized) probability for the classification of that image.

Last updated